Measuring Pedestrian Accessibility: Comparing Central
Business and Commercial Districts in Beijing, London,
and New York City
Summary: This paper proposes a new indicator for urban accessibility - design and
measurement - and introduces a new variable in the evaluation of the urban core. The
connection between accessibility and urban design is established by modeling
pedestrian accessibility from major transport hubs to key points of interest in a
downtown urban environment (specifically areas of employment and commerce). An
investigative study was conducted in 2008 to test this indicator in three major urban
cities: Beijing, London, and New York.
The study reveals impressive data on levels of accessibility and coordination of land use
and public transport service planning, and further sheds light on the impact that urban
compactness, built environment, network design, and pedestrian infrastructure have on
the provision of access to goods and services in a city’s urban core.
Results of the study indicate that pedestrian accessibility is a measure, not only of
infrastructure design and quality, but also of the successful arrangement of urban
resources (i.e., goods, services, recreation, greenery). Further, this paper suggests that
improved urban design will lead to more attractive and effective public transport
systems.
Authors: Mariana Torres (World Economic Forum); Li Yanan (Independent Consultant);
Emily Dubin (Independent Consultant); Shomik Mehndiratta (World Bank)
1
I. Introduction
When on lunch break, how far must you travel from your office to access a restaurant
and how convenient is it to walk there? From the World Bank Office in Beijing, we often
walked for an average of 18 minutes to reach one of our favorite lunch spots in Beijing’s
Central Business District (CBD), which was located a mere 100 meters distance across
the street from our office building. The reason: Beijing’s third ring road, which includes
few pedestrian crossings, stands directly in between the two locations, and practically
cuts through this growing CBD.
Accessibility, when applied to urban transportation, measures the ability and level of
convenience with which people access goods and services. Accessibility is a key
indicator defining a city’s success, given high levels of access provide populations with
the opportunity to connect through marketplaces of goods, activities, and ideas.
Ultimately, providing access to these opportunities is transportation’s main goal, as cities
develop and progress with enhanced relationship-building, exchange, and connectivity.
Three levels of urban transport accessibility are required to achieve high connectivity in a
city: (1) road access for motorized vehicles (e.g. a well-connected road network); (2)
public transport access (e.g. a high coverage system); and (3) non-motorized access (e.g.
safe and convenient pedestrian and bicycle facilities). While emphasis is often placed on
road and public transport infrastructure, pedestrian and non-motorized access is too often
overlooked. Such pedestrian access is particularly valuable in terms of continued growth
and long-term sustainability of urban cores, especially within the fastest-growing cities of
the developing world, which often have a tendency towards sprawl and therefore higher
costs attributed to connectivity.
Giving adequate attention to accessibility is important because appropriate provision of
pedestrian access is critical to the success of a public transport system. When pedestrian
accessibility is reliable, safe and convenient, public transport options become the
preferred mode of transport in urban centers and personal motorized vehicles. This
concept is especially relevant to trip-makers who have a choice between public and
private modes of transport; even extensive networks cannot compensate for a failure in
adequate access to public transport facilities. In countries like China, whose cities follow
a national mandate to give priority to public transport through the implementation of bus
and metro systems, particular attention to pedestrian accessibility is vital to success.
Finally, pedestrian accessibility and walkability is important in and of itself. As
pedestrian access increases, people become less dependent on motorized transit options,
private motorized transport in particular. Many cities in the developing world, and nearly
all cities in China still show that, on average, more than fifty percent of person-trips made
within the urban core are made either by foot or bicycle. By further improving pedestrian
accessibility, there will be less pressure to invest in the large-scale and costly
infrastructure required to accommodate large volumes of motorized transport – public
and private. Furthermore, despite the importance of quality of pedestrian facilities (e.g.
2
proper sidewalks that are safe and involve adequate facilities), the pedestrian network
must lead people to where they need to go.1
An investigative study was performed in 2008 in order to provide insights into levels of
accessibility as well as the coordination of land use and public transport planning. This
was achieved by testing out a basic accessibility indicator: number of jobs and square feet
of commercial floor-space area accessible within a 10 and 20 minute walking radius of a
major public transit station in the city’s central commercial district and central business
district (CBD). Data was gathered in three metropolitan areas: Beijing, London, and New
York, and compared to provide the following results and conclusions offered below. In
the sections that follow, study findings provide insight into the urban design variables
including urban compactness, built environments and pedestrian friendliness (landscapes,
lighting, sidewalk quality, etc.), which have a causal effect on the survey results.
Conclusions on the significance of the different degrees of accessibility and the relevance
of this work follow.
II. Defining an Indicator for Pedestrian Accessibility
Pedestrian accessibility measured through walking surveys were conducted in October
2008 and August 2009 in Beijing, London, and New York City. These surveys modeled
the distance, direction and obstacles encountered in traveling from major transport hubs
to key points of interest in a downtown urban environment. The surveys were based on
catchment areas determined by 10 and 20 minute walks in four cardinal directions from a
major metro station in the central commercial and business district areas of each of the
three cities. Eight total walks were performed for each city, one in each of the four
cardinal directions for each of two transport hubs per city.
The public transport hubs chosen were selected on the basis of their high influx of
passengers and location in important business and commercial districts in each city. In
the case of Beijing, the fast-growing CBD area is centered around Guomao Metro Station,
while Xidan Metro Station is the main station to access the city’s most important
commercial powerhouse. Similarly, Bank Street and Oxford Circle in London are the two
main terminals to access the city’s CBD and the commercial center. New York City’s
Grand Central Terminal, with a daily influx of 125,000 commuters a day, is the most
important public transit terminal in this high-density job area. The 34th Street Herald
Square Metro Station lies at the heart of New York City’s commercial district.
Once the end points of each walk were determined, a catchment area was defined by
creating a circle connecting these points with the transport hub at its center. A visual
representation of these catchment areas for all six locations is displayed in Figure 1 below.
This graphic provides a comparison between the walking distances possible within 10
and 20 minute time ranges in each of the three city environments.
1
This work is meant to complement the Global Walkability Index (Holly Krambeck, 2006), which focuses
on the importance of high quality pedestrian infrastructure, amenities, and services:
http://dspace.mit.edu/handle/1721.1/34409?show=full.
3
Figure 1. Catchment areas for all locations in Beijing, London, and New York City.
NYC
London
Approx miles traveled in North-South
and East-West directions
NYC Average, by Direction
London Average, by Direction
Beijing Average, by Direction
Beijing
20-minute
N-S
0.63
0.81
0.65
E-W
0.72
0.76
0.88
Results shown in Figure 1 suggest that, within the same allotment of time, different
distances, and thus business and commercial services, are reached in each city, indicating
the differences in “walkability” amongst the survey sites in each of the cities. The
4
variable walking lengths, as reported above, are influenced by factors such as levels of
pedestrian congestion, number of intersection and pedestrian light crossing phases, as
well as size and quality of the sidewalks.
In New York City, the average distance in miles covered in 20 minutes was 0.68 miles:
0.63 miles travelled in the north-south direction and 0.72 miles travelled in the east-west
direction. London, on the other hand, allowed for more equal distances traveled between
north-south and east-west directions. In study areas within London, it was found that
pedestrians could travel approximately 0.8 miles in each direction in 20 minutes, a 16%
increase from New York City in distance traveled overall in 20 minutes.
The distance traveled in Beijing largely falls between the New York City and London
distances with an overall average of 0.76 miles in 20 minutes. This is equivalent to 13%
increase in overall distance traveled in 20 minutes in New York City. Interestingly, when
compared to London, Beijing offers shorter pedestrian distances in all directions and all
times except for in the 20 minute walk east-west direction, where 15% and 21% more
distance is travelled compared to London and New York City respectively. These
relationships highlight the mobility differences for pedestrians between the three urban
cores.
When considering distance travelled within a fixed time period as the principal indicator
for pedestrian accessibility in a city, it would be easy to conclude, based on the figures
above, that Beijing is more accessible than London or New York. However, as Figures 2
and 3 below demonstrate, the radius of the catchment area does not coincide jobs and
commercial area accessible – an issue of urban design. This is particularly true for New
York City, where, despite a smaller catchment area, the high level of building density and
urban compactness allows equally high access to jobs and services within a limited area.2
2
The job and commercial floorspace area calculations are based on estimates. Please refer to Annex 1 for
further information on the methodology.
5
Figure 2. Number of Jobs Accessed in 10 min and 20 min
Figure 3. Commercial Areas Accessed in 10 min and 20 min (square meters)
Unit：1,000 square meters
The data provided in Figures 2 and 3 are revealing, as they highlight the variable access
to job and commercial facilities in each studied city. Different levels of accessibility
among the commercial and business centers is seen as particularly high in terms of
commercial floorspace accessible: where New York City boasts access to 20 percent
more area than London, and as much as 97 percent more area than Beijing. In terms of
job access, New York City contains 80 percent more jobs available within the 20 minute
catchment area than does London, and 91 percent more than Beijing. Less of a difference
exists between London and Beijing centers, but this difference is still significant for the
20 minute catchment area, where London contains 85 percent more commercial
floorspace and 55 percent more jobs than Beijing.
6
Results of this data indicate that, for both the commercial and business districts, New
York City is much more pedestrian accessible than either London or Beijing, and that
though the survey areas studied in Beijing allowed the greatest distance for pedestrian
travel, this walkability does not allow for the greatest accessibility for business and
shopping.
III. Analysis
Results of this study yielded interesting data in terms of the variation of distances
travelled in the business and commercial districts studied, and the number of resources
(jobs and commercial floorspace) accessible within the catchment areas. While the
relatively short walkability within New York City’s urban core was surprising, the high
number of jobs and commercial area accessible was consistent with expectations because
of how the City is built – up instead of out. On the other hand, urban design elements
including building compactness, floor-area ratios (FARs), block size, and road network
design contribute to the walkability of Beijing’s urban core compared to London or New
York – indicating an easier, more efficient walking environment – though at the price of
lower accessibility.
The visual representation from a footprint of a Google Earth (2009) aerial imagery
displaying 500x500 meter snapshots of the three CBD catchment areas, as shown in
Figure 4 below (the length of white lines is 500m), indicates very different planning
models and urban design between the three CBDs. Each urban core displays variations in
building height and proximity, road network design and block size, and pedestrian
infrastructure and visual environment. These are all important variables in a city’s level
of accessibility, as discussed below.
Figure 4. Google Earth Aerial View of Urban Density (2009)
NYC
London
Beijing
1) Urban Compactness
Perhaps the most significant and variable factor between survey areas of the three cities is
the compactness of the cities’ built area, especially the height and proximity of buildings
within the urban cores. Within the catchment areas, New York features buildings that are
both tall and built near each other, which allows for a large number of jobs and services
within a smaller area. The situation is slightly different in London, where buildings are
7
not as tall – an average of five stories in the commercial catchment area – but they are
built close together.
Urban design in Beijing is nearly opposed to what is observed in New York: despite the
presence of a significant number of high-rise buildings, the wide layout is spaced out,
resulting in low horizontal capacity for jobs and commercial floorspace area. This is due
in part to Beijing’s urban construction regulations, which stipulate a much wider area to
be left in between buildings and between buildings and the edge of the road when
compared to New York or London. Figure 5 below illustrates the differences in layout
between New York and Beijing: New York’s densely built design and the underutilized
spacing in Beijing have a significant impact on the number of jobs and commercial
floorspace available.
Figure 5. Comparing urban density in commercial cores in New York City (34 Street Herald
Square) versus Beijing (Guomao)
NYC
Beijing
Further, Chart 1 below shows the approximate number of buildings present within the 20minute catchment area of each business and commercial district studied. As expected, the
number of buildings in New York City is significantly higher: 70% more buildings than
Beijing, but only 40% more than London.
City
Approx. Number of Buildings in “20-min circle"
NYC
London
Beijing
710
430
220
Chart 1. Building Density Information of three cities (Mean Value of Two Target Area)
8
2) Road Network Design: Qualitative Differences in Secondary Road Network, Block
Size, and Community Design
Crucial to urban compactness is the layout of the road network. New York City features a
dense urban grid with small blocks and narrow roadways, in other words a solid
secondary road network that allows flexibility in accessing different addresses. Beijing’s
road network, on the other hand, is marked by wide avenues and few, narrow secondary
roads. Figure 6 displays the road network and differences in block size in the research
area, confirming that the road network density in Beijing is clearly lower than that of
NYC.
Figure 6. Road Network Format in New York City versus Beijing
*Image capture from Google Earth: 2010.
NYC: Grand Central Station
Beijing: Guomao Metro Station
The catchment areas studied present differences both the width of the roads and the
length of the road network. Chart 2 below summarizes the differences in average road
width and total network length between New York City, London, and Beijing for the
catchment areas. New York City’s roads are 29 percent narrower than Beijing’s and 34
percent wider than London’s. However, in terms of total road length, New York City’s
network is 52 percent longer than Beijing’s and 15 percent longer than London’s. While
road width may result in higher allocation of road space for pedestrians and therefore a
more comfortable displacement, it provides less flexibility for more convenient access
throughout the catchment area than does a longer road network.
9
City
Average Road
Width (feet)
Approximate Total Road
Length (miles)
NYC
London
Beijing
66
43
98
21
17
10
Chart 2. Road Information of NYC, London and Beijing (Mean Value of Both Catchment Areas)
The main difference in road network design between the catchment areas in the three
cities is block size. Higher road network density in New York City is a product of the
presence of more but smaller blocks compared to Beijing and London, as shown in Chart
3 below. Beijing’s lower road network density and limited secondary network is a result
of the city’s widespread “superblocks”. In fact, Beijing’s superblocks within the
catchment areas are, on average, almost four times larger than New York City’s, and
therefore Beijing’s catchment area contains less than half the number of blocks than New
York City’s. London lies in between NYC and Beijing, with an average block size 60
percent larger than New York’s, but 53 percent smaller than Beijing’s, and nearly
doubles Beijing in number of blocks. Block size is relevant to accessibility to the extent
that it impacts the time it takes pedestrians to cross from one side of the block to another
(to access, for example, a main road or a public transport station), and the amount of
commercial and work places available within a determined area, since these are often on
the edges of the blocks.
City
Average Size (feet*feet)
NYC
London
Beijing
656*197
689*312
853*590
Number of Blocks in 20-min
catchment area
178
143
84
Chart 3. Block Information of the three cities (Mean Value of the CBD and Commercial Areas)
The relevance of the block size is compounded by the micro design of the actual block
areas: land use and building distribution. NYC tends to use the area within its smaller
blocks more efficiently than Beijing: buildings are built closer to each other and nearer to
the road and sidewalk, taking maximum advantage of the area available. Beijing, on the
other hand, tends to distribute constructions in its superblocks according to planning of
gated community areas, whether commercial or residential. The design of these gated
communities hinders free access throughout the superblock. Figure 7 below illustrates
this situation: path 1 and 2 drawn below are the ideal routes for pedestrian travel from
Xian Metro station (the yellow placemark west of the block) to a typical residential
building (identified in the figure by a green house). However, the gated community
design requires that pedestrians use Path 3, equivalent to 50% additional walking time.
Compared to Beijing, London has no gated complexes. Instead, because London’s
original design was based on narrow roads for pedestrians, the equivalent area of a
Beijing superblock in London would generally involve a series of narrow roads traversing
the block and therefore allowing higher access to the premises.
10
Figure 7. Pedestrian Route in Beijing
Path 1
Path 3
Path 2
An additional factor impacts pedestrian accesibility: traffic light crossings. The variation
in traffic signal cycle times is very high for the three cities, as shown on chart 4 below.
By far, New York City has the most predictable and short waiting times with uniform 90
second cycle times. Beijing has the longest waiting times – nearly 60% higher than New
York City and 40% higher than London –, creating significant delays for pedestrians.
Beijing also has the lowest level of predictability, given the high levels of variation in
cycle times across the different intersections (265 seconds vs. 91 seconds for London).
As in other comparisons, London is a middle point between New York and Beijing
averages.
City
Cycle Time
(minutes)
Level of Variation within
Survey Area
NYC
London
Beijing
1:30
2:15 (average)
3:40 (average)
None – fixed timing
Varies based on time and location
Varies based on location
Chart 4. Road Information of NYC, London and Beijing (Mean Value of Both Catchment Areas)
3) Pedestrian Infrastructure and Visual Environment:
Safety and comfort are critical to the perception of the pedestrian’s walking experience,
as is the visual content of the walk in terms of commercial activity, greenery, etc. The
walking experiences in Beijing, London, and New York are very different. The
snapshots below demonstrate some of the differences in terms of visual environment,
obstructions, pedestrian congestion, and width of sidewalks.
11
Figure 9. Pedestrian environment and visual landscape in New York, London and Beijing
New York
City
London
Beijing
IV. Analysis of Results
Identifying an adequate measure for accessibility in a city is important because it is a
measure of how well a city performs bringing people together and providing access to
services. This simple indicator for accessibility, based on number of jobs and commercial
floorspace area available within a 10 and 20 minute walking circumference, illustrates the
“accessibility” component of walkability and highlights key urban design elements which
enable such access, therefore establishing the impact that urban design has on developing
a sustainable urban transport system.
The study demonstrated that a city with design characteristics like New York that gives
way to a much more accessible urban environment than a city with the design
characteristics of Beijing: fewer, wider roads, large superblocks and widely spaced
buildings set back from the road. The study brought to light the importance of the
following considerations when designing a detailed urban area plan for high accessibility:
1) High density land use, a closely built set of high-rises;
2) Dense grid network with sufficient secondary roads and small blocks;
3) Pedestrian-friendly environment.
The indicator proposed in this paper, assessing how people are able to access jobs,
services, recreation, and other points of interest on foot, is therefore useful in planning for
improved urban design in the central business district of major cities as this can partially
remove the burden on a city’s public transport and road network system, and help to
achieve a more efficient urban transport system. In addition, a more sustainable
individual travel pattern also means significant time and economic savings, both for the
end-user as well as for the city’s investments in public transport and road infrastructure.
Further, and crucial to megacities like Beijing, London, and New York, pedestrian
accessibility is key to the success of a public transport system: no matter the number of
metro lines or the quality of the system, if pedestrians cannot access the stations from
their point of origin, they will opt for alternative modes of transport.
The accessibility indicator proposed in this paper can be used across cities as a key tool to
help decision makers understand how their urban cores compare and plan for enhanced
walkability and general service access in urban environments, including improved
12
utilization of public transport options and pedestrian access. Further, this indicator should
form part of a standard set of urban performance indicators, given its importance to cities
in terms of determining productivity and linking to sustainability in urban areas through
higher connectivity and therefore increased non-motorized access. Finally, particularly
for the case of China’s fast-developing cities, identifying and implementing the urban
design variables that are conducive to more people-oriented cities will enhance
competitiveness to the country’s new and developing urban communities.
13
Annex
 Annex 1: Study Methodology and Assumptions
 Annex 2: Presentation about Detailed Survey Description
 Annex 3: Walking Video of NYC, London and Beijing
For More Information
Website Link: http://go.worldbank.org/OWNNK85QO0
Contact: Shomik Raj Mehndiratta (smehndiratta @worldbank.org), East Asia and Pacific Region Transport
14
Annex 1: Calculation Methodology and Assumptions
WALK
Methodology
•
•
•
•
•
Not performed during rush hour
Walk in a constant rate
Continue in one direction without unnecessary
street crossings
Not walk during red lights
If a pedestrian reaches a dead end that cannot
be avoided or walked around (such as a
river..), the pedestrian will stop.
Assumptions
•
•
•
•
Workers use the metro exit closest
to their office
No accidents or delays en route
No road closures or construction
activities removing sidewalk access
Constant walking rate
15
ANALYSIS/PROCESSING
Methodology
•
Jobs and commercial space reported every
minute are sum totals for that minute in all
directions, not just the direction walked (see
Figure 1 below)
Figure 1. Area of cumulative analysis by minute.
Commercial space and jobs were calculated by
minute total in all directions, rather than by cardinal
direction (the light blue area rather than the red
hatched area)
Assumptions
•
•
•
•
•
•
•
•
Total commercial space and jobs in London
and New York City were calculated assuming
1 employee for every 300 square feet of
commercial space, as per the DeChiara
Planning & Design Criteria Handbook .
Because of different planning requirements
and context, this calculation was not used for
Beijing.
Figures used in this study were rounded to the
100's place in order to avoid giving a false
sense of accuracy (as these figures are
estimates, meant only to provide a general
framework for comparison between cities).
All commercial area was converted to meters
for easy comparison
•
•
For this study, we had to assume
that our data was generally up-todate, or that figures hadn't changed
drastically since the last reporting
period.
If the study area (circle) created by a
walk cut through a block, lot or
building, total commercial area or
jobs would be taken proportionally
(ie. if 1/3 of a building is within the
10 minute circle, only 1/3 of the
building's commercial space is
accessible).
Assumptions were made based on
the number of jobs per commercial
area, and visa versa, for both
London and New York City (see
Methodology, to the left).
For New York City, we assumed that
any taxlot coded for Commercial and
Office space were commercial areas
with employees, and that employees
were located throughout the building
- one every 300 square feet.
For London, we assumed that for
every job per block, there were 300
square feet of commercial space
surrounding the employee.
For Beijing jobs, we assumed that
the employment numbers is equally
distribution by area in one traffic
zone.
For Beijing commercial areas, we
assumed that for each shopping
mall, all the area that belongs to this
building is commercial area and no
office area.
16